The meeting where the numbers stopped making sense
"Explain this to me like I'm wrong."
The CFO wasn't angry. He looked confused — which, in a boardroom, is usually worse.
"We froze hiring. We reduced discretionary spend. We're within budget. So why are margins down?"
No one had a clean answer. Not because the data wasn't there, but because the data wasn't telling the right story. HR reported stable headcount. Finance confirmed cost controls within 1.8%. Operations argued productivity was up. Each team was correct in isolation. Collectively, they had missed the actual question.
This was a post-pandemic restructuring at one of the larger Latin American airlines. Margins had dropped 4.2 points in two quarters despite flawless budget adherence. The model itself was broken.
The pattern is not unique to aviation. Weeks earlier, in a separate boardroom with a global retailer, the CEO had frozen hiring to protect margin based on financial reports. Within months, store performance slipped. Service levels fell as experienced staff burned out and newer hires struggled to close the gap. The decision looked sound on paper. The hidden costs — lost expertise, disengagement, and erosion of customer experience — only surfaced when NPS started to fall.
Different industries. Same failure mode. Labor cost has outpaced the way most organizations measure it.
Three questions that redefine the problem
Every CFO I have worked with in the past eighteen months asks some version of three uncomfortable questions:
What are we actually measuring when we say "labor cost"?
How is AI structurally breaking the traditional workforce economics?
Why do current workforce planning models fail executives — and what architecture must replace them?
Let's take them one by one.
1. What are we actually measuring when we say "labor cost"?
In most organizations, labor cost is still treated as a static accounting figure: salary, benefits, taxes, fully loaded cost per FTE. Clean. Structured. Completely insufficient.
Because labor is not static. It moves, leaks, compounds, and — more dangerously — misallocates.
Labor often represents more than 50% of operating expenses. The real issue is not magnitude. It is opacity. What typically stays invisible:
The cost of unfilled roles (vacancy drag)
The cost of misaligned skills (productivity decay)
The cost of turnover (replacement plus ramp inefficiency)
The cost of poor allocation (right people, wrong work)
The cost of not adopting AI in core processes
Recent industry analysis suggests that hidden costs from workforce misalignment can erase up to 15% of potential annual productivity gains in large organizations, and that failure to embed AI in core functions incurs opportunity costs of 5–12% of departmental operating budgets. These are not accounting losses. They are strategic losses hidden in plain sight.
The Latin American context sharpens this. In markets such as Mexico, Colombia, and Brazil, severance and statutory costs can reach 30–40% of annual compensation. A poorly executed hiring freeze in these jurisdictions does not save money — it creates a deferred P&L event the CFO will discover eighteen months later, usually in a termination wave triggered by the next reorg.
The airline case illustrates this precisely. A "cost-saving" hiring freeze reduced payroll by 9% in the first year. Within six months, project delays had increased by 23%, overtime had risen by 17%, and voluntary attrition had spiked by 8 points. Net effect: margin erosion, not margin protection.
The CFO later summarized it better than any model: "We didn't reduce labor cost. We redistributed it badly."
That distinction is everything. And it explains why some organizations succeed quietly where others fail loudly. BGIS, a facilities management firm, didn't reduce headcount to protect margins — it reduced voluntary attrition from 25% to 15% over two years by rebuilding how it measured workforce cost. That single shift in measurement logic was worth more than a decade of hiring freezes.
2. How AI is structurally breaking the equation
For decades, the implicit formula was simple:
Labor cost = cost per employee × number of employees
AI destroys that simplicity. Not gradually. Structurally.
The new reality looks closer to this:
Labor cost = cost per capability × (human effort + AI augmentation)
That shift introduces three non-linear effects.
First: capability replaces headcount as the unit of value. Two employees with identical salaries can now produce radically different outputs depending on the AI tools used, data accessibility, and process integration. The variance is no longer marginal. In client work across financial services, we have seen it exceed 40–60% within the same job family, grade, and compensation band.
Second: productivity gains are no longer predictable. In traditional planning, improvements were incremental. With AI, they are discontinuous. A single AI-assisted workflow redesign can eliminate entire task layers while increasing throughput. Most organizations capture the upside without recalibrating their labor cost models. The gains are real. The economics remain misunderstood.
Third: cost structures become layered and interdependent. Labor is no longer just labor. It is human cost plus AI tooling cost plus data infrastructure plus governance overhead. In most P&Ls, these sit in completely different categories — HR owns labor, IT owns tooling, FP&A owns the financial impact. Disconnected in planning, intertwined in reality. Which leads to flawed decisions, and to CFOs who cannot explain why their cost per unit of output keeps rising while headcount falls.
The fix is not technological. It is organizational. Leadership teams that bridge this gap do three things consistently: they run joint planning sessions between HR, IT, and Finance; they set shared KPIs that measure workforce effectiveness and technology adoption together; and they treat workforce planning as a cross-functional discipline rather than an HR cycle.
In a capability-driven P&L, the most expensive employee is no longer the one with the highest salary. It is the one whose work an AI could do in half the time — while your planning model still codes them as fully productive.
3. Why workforce planning fails — and the architecture that must replace it
Most workforce planning initiatives do not fail technically. They fail conceptually. They produce outputs, but not insight.
The three systemic failures
HR plans people. Finance plans cost. Business plans growth. Three plans, three logics, one inevitable misalignment. When the CFO asks how a product launch affects three-year capability costs, no one owns the answer, because no single model connects demand drivers, skills supply, and financial impact.
Time horizons do not match either. HR thinks in hiring cycles. Finance in fiscal periods. Business in market windows. Each function is correct in isolation and misaligned in aggregate.
And the model is backward-looking. It explains what happened. It does not simulate what could happen. So when executives ask the questions that actually matter — What happens if we automate 20% of this function? What is the cost of not hiring this role? What is the financial impact of delayed capability? — the system breaks. Not because it lacks data. Because it lacks integration.
The two planes that have to merge
A working workforce planning process unites two planes that most organizations still run on separate infrastructure.
The strategic plane: what capabilities and skills does strategy require, when, and with what risk if gaps remain open?
The financial-operational plane: what does each decision cost, by entity, cost center, job family, and scenario?
Enterprise Performance Management platforms — Oracle EPM being the most complete example for this class of problem — articulate both planes by combining Strategic Workforce Planning with Workforce Planning and integrating to HCM and ERP. The technology matters. But the value is architectural, not functional.
MIT Technology Review Insights research suggests organizations that integrate workforce and financial planning report up to a 20% improvement in decision speed and quality versus those operating in silos. The improvement is not about faster spreadsheets. It is about a shared model of reality.
What changes when it works
The conversation shifts. From "How many people can we afford?" to "What combination of capabilities delivers the best economic outcome under different scenarios?"
Published case benchmarks make this concrete:
KPMG reduced forecast days by 33% and cut presentation preparation time by 35% after consolidating roughly 200 manual templates into a unified planning model.
Pearson collapsed budget revision cycles from more than three days to twelve hours, with payroll processing effort down 20%.
Securitas accelerated hiring speed by 70% and compressed monthly reporting from three weeks to two minutes — after previously operating across 27 disconnected platforms.
BGIS, as noted, reduced voluntary attrition from 25% to 15% over two years, a shift worth more than any cost-containment program of comparable period.
These cases carry publication bias and should be read as directional, not as guaranteed outcomes. But the pattern is consistent: when HR and Finance stop reconciling and start co-modeling, the planning cycle shortens, the data becomes defensible, and the executive conversation moves from cost debates to capability decisions.
That shift is what makes workforce planning strategic. Not the technology. The integration.
Closing: the question that changes everything
At the end of that airline meeting, the CFO did not ask for another report. He asked a harder question:
"Can we model the real cost of how work actually gets done in this company?"
Not just salaries. Not just headcount. The full system.
That question changes the whole conversation. Because the real problem is not a data problem. It is a model problem.
Three ideas to distill:
Labor cost is not an accounting figure. It is a capability equation, and the unit of analysis has changed.
AI does not just improve productivity. It restructures the economic grammar of the organization, and most planning models have not caught up.
EPM architecture is not about automation. It is about forcing HR, Finance, and business onto a single decision model — scenario-ready and auditable.
This week, three questions to ask your team:
What percentage of our labor cost line is visible by capability, not by cost center?
Can HR, FP&A, and one business leader reconcile headcount definitions in a single one-hour meeting, with one reconciled number at the end?
What is our "cost of not hiring" calculation for the three most critical open roles on our plan today?
If you cannot answer any of these cleanly, you do not have a workforce planning process. You have a payroll budget.
The 12 to 24 month horizon. Organizations that rebuild their workforce costing models on EPM-class architectures will operate on a different economic grammar than the rest. Published case benchmarks point to payback windows of 12–24 months and three-year ROI ranges of 80–250% (estimated, with meaningful variance by sector, data maturity, and implementation governance). The gap between leaders and laggards is not technological. It is how quickly leadership teams accept that workforce planning has moved from an HR cycle to a margin-protection discipline.
The companies that answer that will not just improve workforce planning. They will redefine performance itself.
The rest will keep managing headcount.
And wondering — quietly — why the numbers don't add up.
Pedro San Martín
Principal — Asher & Company
